文档介绍：第 32 卷第 1 期计算机应用研究 Vol. 32 No. 1
2015 年 1 月 Application Research puters Jan. 2015
基于分布式粗粒度并行计算的遗传规划算法研究*
李志坚,吴晓军,任哲坡,欧小波
( 西北工业大学自动化学院,西安 710072)
摘要�针对经典遗传规划算法( CGP) 存在容易早熟收敛、运行效率低的缺陷,提出一种将分布式计算与遗传
规划算法结合的计算模型。该模型利用个体迁移策略实现对种群的优化,克服易早熟的缺陷。并且采用分布式
计算能够有效地节省算法的运行时间。最后通过对语音数据预测误差的比较,验证了改进后算法的有效性。实
验表明,基于分布式粗粒度并行计算的遗传规划算法( CGGP) 计算性能优于经典遗传规划算法( CGP) 。
关键词�遗传规划; 子种群; 收敛; 分布式计算
中图分类号� TP181; TP301. 6 文献标志码� A 文章编号� 1001-3695( 2015) 01-0048-03
doi�10. 3969 /j. issn. 1001-3695. 2015. 01. 011
Research of ic programming based on
distributed coarse-grained puting
, , ,
LI Zhi-jian WU Xiao-jun REN Zhe-po OU Xiao-bo
�, , , �
School of Automation Northwestern Polytechnical University Xi’an 710072 China
���,
Abstract Based on the deficiencies of classical ic programming CGP such as convergencing prematurely and operat-
,
ing efficiency this paper proposed puting model bined puting with ic programming. The
, ,
model optimized populations through the individual migration strategy overcame the deficiency of prematurity and saved the
,
running time of the algorithm effectively through puting. Finally it validated the effectiveness of algorithm
which improved paring the prediction error of the speech data. The experiment demonstrated that the calculated
��
performance of the ic programming algorithm based on distributed coarse-grained putin